Zhuo Deng

Papers

Total Citations

H-Index

About

Zhuo Deng is a computer vision researcher whose work focuses on advancing human behavior understanding from visual data, particularly in challenging real-world scenarios. His most notable contribution is the development of MEBOW (Monocular Estimation of Body Orientation in the Wild), a seminal method for estimating human body orientation from single images. This work addresses a critical gap in applications like robotics and autonomous driving, where traditional 3D pose estimation fails due to poor resolution, occlusion, or indistinguishable body parts. To support this research, Deng created the COCO-MEBOW dataset, providing a large-scale benchmark that has become a standard resource in the field. His paper on MEBOW has garnered over 40 citations, reflecting its impact on enabling robust visual perception under difficult conditions. By focusing on body orientation—a simpler yet highly informative cue—Deng has provided a practical alternative for systems that cannot rely on full pose estimation. His work continues to influence the development of more resilient and efficient computer vision systems for real-world deployment.